Unclassified PUMA/RD(99)4/FINAL

Organisation de Coopération et de Développement Economiques OLIS : 13-Dec-1999Organisation for Economic Co-operation and Development Dist. : 14-Dec-1999__________________________________________________________________________________________

Or. Eng.PUBLIC MANAGEMENT SERVICEPUBLIC MANAGEMENT COMMITTEE

MANAGING ACCOUNTABILITY IN INTERGOVERNMENTAL PARTNERSHIPS

Unclassified

PU

MA

/RD

(99)4/FIN

AL

Or. E

ng.

For further information, please contact Elke Loeffler,Tel: (33-1) 45 24 90 76, E-mail: elke.loeffler@oecd.org

85517

Document complet disponible sur OLIS dans son format d’origine

Complete document available on OLIS in its original format

PUMA/RD(99)4/FINAL

2

This report and the related case studies are available on internet: http//:www.oecd.org/puma/

Disponible également en français sous le titre :

L’Affectation des Responsabilités dans les Partenariats entre Niveaux d’Administration

© OECD, 1999Applications for permission to reproduce or translateall or part of this material should be made to:Public Affairs and Communications, OECD,2, rue André-Pascal, 75775 Paris Cedex 16, France

PUMA/RD(99)4/FINAL

3

FOREWORD

The trend toward administrative devolution in many OECD Member countries requires a focus on agreedresults-oriented performance in order to assure accountability. Performance partnerships promise to be aneffective tool to address shared accountability of intergovernmental projects and programmes.

As part of its work in intergovernmental management, the Public Management Service has studiedintergovernmental partnerships in OECD Member countries in order to identify the key issues andpractices to improve accountability for shared responsibility. The report was prepared by Elke Löffler ofthe OECD Public Management Service. Technical assistance was provided by Hélène Leconte.

The views expressed are those of the author and do not commit or necessarily reflect those of governmentsof OECD Member countries. This report is published on the responsibility of the Secretary-General of theOECD.

PUMA/RD(99)4/FINAL

4

TABLE OF CONTENTS

EXECUTIVE SUMMARY 5

INTRODUCTION 7

PART I: THE NEW RELEVANCE OF INTERGOVERNMENTAL PARTNERSHIPS

AND THEIR IMPLICATIONS FOR ACCOUNTABILITY MANAGEMENT 9

Devolution supporting the formation of partnerships between levels of government 9

Definitions and focus of the study 12

The analytical framework 16

The concept of performance partnerships 17

PART II: ACCOUNTABILITY MANAGEMENT IN DIFFERENT KINDS

OF INTERGOVERNMENTAL PARTNERSHIPS 20

Types of intergovernmental partnerships 20

Accountability management of “joint production” partnerships 22

Accountability management of “common resource” partnerships 24

Accountability management of “joint investment” partnerships 25

PART III: PERSPECTIVES AND LIMITS OF INTERGOVERNMENTAL 27PERFORMANCE PARTNERSHIPS

REFERENCES 28

Notes 30

PUMA/RD(99)4/FINAL

5

MANAGING ACCOUNTABILITY IN INTERGOVERNMENTAL PARTNERSHIPS1

EXECUTIVE SUMMARY

The study explores the nature of different kinds of project and programme partnerships between levels ofgovernment and considers the strengths and weaknesses of partnership institutions such as special purposeagencies or commissions in addressing issues of accountability. Due to the forces of devolution andglobalisation, many OECD Member countries have shifted tasks downward to provincial and localgovernments, but also upward to international organisations and similar fora. This delegation ofresponsibility may take different forms, depending on whether it is policy, management or financialresponsibility which is transferred to another level of government. The report mainly focuses on jointadministration of programmes or projects between different levels of government, which implies thataccountability has to be shared between levels of government. This kind of intergovernmental partnershipposes major management challenges with regard to the development of new accountability frameworks andperformance measurement.

Partnership arrangements between different levels of government have to consider three kinds ofaccountability relationships:

• accountability among the partners;

• accountability between each partner and its own governing body; and

• accountability to the public.

As the case studies show, traditional forms of administrative co-operation are mainly cost-sharingarrangements with an inherent risk of diluting or diffusing accountability. The lack of performanceaccountability stems from a missing specification of objectives or evaluations of the partnerships. Theanalysis of different kinds of intergovernmental partnerships points out that the nature of the accountabilityproblems varies with the type of project partnership involved:

• The so-called “joint production” partnerships focus on the provision of technical services such asservices to property, and the management of general infrastructure such as waste collection, watersewage and the operation of other joint utilities. Due to their productive nature, it is not difficult tospecify output targets. But since each member only has to cover a certain part of the total costs,the members of the partnership have little incentive to evaluate the overall performance of thepartnership.

• The so-called “common resource” partnerships deal with the management of common resourcessuch as natural resources and land. Partnerships for the sustainable development of rivers and othercommon goods often do not define outcome targets.

• The so-called “joint investment” partnerships are intergovernmental investment projects for thecreation of infrastructure or production facilities. Large-scale investment projects in the publicsector generally do not have performance targets. Due to the time frames and the multitude ofstakeholders involved in investment projects, risk identification and risk allocation tend to be weakas well.

PUMA/RD(99)4/FINAL

6

The concept of performance partnerships between different levels of government allows for bothaccountability of results and for partnership accountability. Performance partnerships are based on thefollowing elements:

• clarity of roles and responsibilities, including the identification of risks;

• joint specification of performance expectations, including balancing expectations and capacities;

• reporting of valid and reliable information on what was achieved among the partners, to the publicand to the respective governing bodies;

• joint evaluations on how far pre-determined overall objectives have been met; and

• feedback on the performance achieved and adjustments.

As the partnership case studies show, the introduction of this form of performance contracting betweenlevels of government has its own challenges. For example, the reporting requirement implies collectingdata from different jurisdictions or levels of government. In some cases, obstacles in the form of privacy ordata protection laws have to be overcome. In other cases, the recipient of the funds may simply tell thenational government to go away and refuse to give data.

The main recommendations to improve the accountability management of specific types of partnershipinstitutions can be summarised as follows:

• “Joint production” partnerships: The establishment of performance contracts between thepartnership institution and the owner governments may ensure that periodic evaluations of thepartnership’s activities take place.

• “Common resource” partnerships: Introducing outcome-based performance contracts may help thepartnership institution shift away from purely compliance-oriented management towards a moreflexible outcome-based management of the common resource.

• Joint investment” partnerships: Project planning and development have to include risk analysis andthe definition of performance goals, and establish evaluation mechanisms early on.

In all three cases, intergovernmental partnerships have to become more transparent to the public.

In many areas of administrative decentralisation, performance partnerships have become an emergingpractice in OECD Member countries, and the report further advocates the use of performance partnershipsto manage joint projects or programmes. Intergovernmental partnerships that involve policies rather thansingle projects or specific programmes have much more complex relationships. This kind of politicaldevolution is beyond the scope of this report.

PUMA/RD(99)4/FINAL

7

INTRODUCTION

The purpose of the paper is assist decision-makers in the public sector to develop new accountabilityframeworks for different types of project or programme partnerships, which are jointly managed byseveral jurisdictions.

The focus is on co-operative partnerships between different jurisdictions as opposed to purely consultativeor co-ordinating types of partnerships. Jurisdictions co-operate if they jointly produce goods or services, ifthey jointly manage the use of public goods or if they pool their resources for investment purposes. Jointproject or programme management between levels of government will usually imply the establishment ofjoint institutions such as consortia or working groups which are at the centre of this study.

The analysis mainly draws from partnership case studies that emerged from an expert meeting at OECD.The approach used for the production of the case studies was innovative, using the perspectives ofrepresentatives of various levels of government. A “neutral” expert with experience in the field put the casestudy in a broader context and assured a balanced assessment of the partnership.

The case studies deal with the following aspects of accountability management in intergovernmentalpartnerships:

a) Basic description of the public sector partnership: the contents of the collaboration, the main actorswithin the partnership, the basic structures and the perceived motivating factor leading to thisparticular partnership.

b) Accountability expectations: description of perceived accountability relationships (individual level,corporate level), of external stakeholders of accountability (parliament, audit court, etc.), of thetypes of performance criteria that are used in various accountability relationships and how they aredefined, and of potential problems in terms of conflicting expectations.

c) Information management: relevance, accuracy, timeliness of information provided by eachpartnership institution to its stakeholders, including the public.

d) Evaluation of information: analysis of individual/institutional actors involved in evaluations of thepartnership institution, of methodology (hierarchical, co-operative, self-assessing) and of process(internal or external, ex-ante or ex-post).

e) Incentive mechanisms: types of incentive mechanisms used (positive or negative, compliance orself-steering), feedback to the partnership institution, perception and response of the partnershipinstitution to the feedback.

f) Uncertainty: does the partnership institution provide for unforeseen events? What kind ofunforeseen events have there been in the past and how have they been managed?

g) Future of the partnership: will the partnership be sustainable? Who will be the future actors andwhat will be the structures of the partnership? What are the lessons learnt?

h) The accountability context of the partnership: pattern of multiple accountability relationships in thegiven country (hierarchical, legal, professional and political). Is the accountability managementwithin the partnership typical of similar public sector partnerships in the country (examples fromother policy areas)? Is there a trend towards performance-based accountability in the givencountry, and if yes, how is it integrated into the more traditional pattern of accountability?

A list of the case studies is provided in Table 1. Each of the case studies and a background paper on legalaspects of intergovernmental partnerships can be accessed via the internet at http://www.oecd.org/puma

PUMA/RD(99)4/FINAL

8

Table 1

Case Studies on Intergovernmental Partnerships

Switzerland The Devolution of Education Policies from the FederalGovernment to the Cantons in Switzerland

Sweden The Devolution of Labour Market Programmes from theCentral Government to the Local Level in Sweden

United States The National Environmental Performance Partnership System(NEPPS) between the States and the U.S. EnvironmentalProtection Agency

Spain Intergovernmental Partnerships at the Local Level in Spain:Mancomunidades and Consortia in a Comparative Perspective

Czech Republic, Germany and Poland Administrative Co-operation in the Euroregion Neisse-Nisa-Nysa

The paper is divided into three sections:

• Part I outlines the new relevance of intergovernmental partnerships and the implications foraccountability management.

• Part II specifies different types of intergovernmental partnerships and analyses how the partneringarrangements between levels of government address issues of accountability. Particular attentionwill be given to performance measurement and evaluation problems of the various kinds of projectand programme partnerships.

• Part III considers briefly the prospects and limits of intergovernmental performance partnerships.

PUMA/RD(99)4/FINAL

9

PART I: THE NEW RELEVANCE OF INTERGOVERNMENTAL PARTNERSHIPS ANDTHEIR IMPLICATIONS FOR ACCOUNTABILITY MANAGEMENT

Devolution2 supporting the formation of partnerships between levels of government

The current trend of devolution in many OECD Member countries gives the traditional debate oncentralisation and decentralisation a new momentum and quality. Changed environments raise again theold question of fiscal federalism: which distribution of responsibilities conforms with the principle ofsubsidiarity and is in balance with the distribution of revenue and expenditure capacities? But apart fromthe classical fiscal “who should do what” issue, there are also the public management issues concerning theappropriate accountability structures and instruments to manage different allocations of tasks betweendifferent levels of government.

Allocational shifts between levels of government may take various forms. At present, two major trends canbe identified: on the one hand, there is a trend towards a clear separation of policy and managementresponsibility, including spending and taxing powers between levels of government. On the other hand,there is also a trend towards shared responsibilities between levels of government. In the latter case, centralgovernments have shifted the implementation of policies to lower or higher levels of government.Nevertheless, the central government may still retain formal responsibility and accountability for thequantity and quality of the services to be delivered at other levels of government, which implies thataccountability has to be shared between levels of government. This situation is at the centre of the report.

Traditionally, the Constitutions of federal countries such as Austria, Germany and Switzerland require asubstantial portion of federal law to be implemented by the States. This kind of administrativedecentralisation now takes place in other OECD Member countries, where the national government passesmanagement responsibilities to subnational levels of government. In many cases, the primary tool tomanage this kind of devolution is partnership arrangements between different levels of government. Thesepartnership arrangements often establish new institutions in order to co-manage programmes or projects.

Example

A case in point for the new relevance of partnership arrangements is the devolution of policy andmanagement responsibilities from the national government to the Autonomous Communities in Spain. Thenational government and regional authorities established bilateral commissions, on a temporary basis, inorder to guide and monitor the transfer of functions negotiated in the devolution process. Once the transferhad been accomplished, however, the commissions continued to operate. These bodies now supervise thenew “Co-operation Agreements” (Convenios de Colaboración) that deal with welfare policies such aseducation, labour, culture, health and social services (see Font, Gutiérrez Suárez and Parrado Díez, 1999).

PUMA/RD(99)4/FINAL

10

Figure 1: “Co-operation Agreements” between the Spanish National Government andAutonomous Communities from 1988-1997 (in total numbers)

176

260

180

261

294316

350

314

389365

0

50

100

150

200

250

300

350

400

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

Source: Font, Gutiérrez Suárez and Parrado Díez, 1999.

Intergovernmental partnerships that are based on shared accountability between several jurisdictionsrequire a new concept of accountability because of their inherent risk of diffusing accountability among theparties involved. There are many indications of failures of traditional forms of partnerships. For example,as an empirical study on partnerships for the development of investment projects found out, cost over-runsof 50 to 100 per cent in fixed costs are common. Even over-runs above 100 per cent like in the case of theChannel Tunnel between France and the United Kingdom are not uncommon (Bruzelius, Flyvbjerg andRothengatter, 1998:424).

The increasing awareness of accountability problems in traditional intergovernmental partnerships isreflected in the fact that there is stronger and stronger criticism about blurred accountability:

PUMA/RD(99)4/FINAL

11

Examples

The German “Lean State” Advisory Council recommended reducing the system of joint financing as itblurs accountability and necessitates a great deal of administrative effort (Geschäftsstelle desSachverständigenrats “Schlanker Staat”, 1997:30). This recommendation referred to government activitiessuch as housing, new universities and economic promotion which became joint tasks of the Federation andthe Länder with the amendment of Art. 91a and Art. 91b in the German constitution in 1969.

The same line of criticism can be found in the Decentralisation Promotion Plan which was adopted by theJapanese Government on 29 May 1998. It stipulates that an excessive co-operation and fusion of centraland local government functions have undermined an efficient and effective performance of the publicsector (Furukawa, 1998).

Accountability problems also exist for municipal service provision in cases where municipalitiestransferred responsibilities to single and/or multi-purpose agencies. So far, however, empirical studies arestill lacking that evaluate the efficiency of service delivery through intermunicipal partnerships.

Example

The Netherlands is an illustrative case because its Act on Intermunicipal Partnerships (WetGemeenschappelijke Regelingen [WGR]) has stimulated the establishment of many such arrangements.Although many services are provided via such pre-selected partner WGR partnerships on the basis ofassumed economies of scale, their efficiency has not yet been analysed. However, serious accountabilityproblems give rise to the assumption that such intercommunal arrangements are a relatively inefficient wayof municipal service provision (see in detail, Boorsma and de Vries, 1998): The board members of theintermunicipal authority will normally consist of some aldermen from participating cities who have noincentive to monitor performance. The members of the council of a partner city will not control theautonomous intermunicipal board either. They only have to pay for part of the total outlays of theintermunicipal arrangement, which have to be weighted against the benefits of services that are spread overall the participating cities. The consequence of the weak accountability structure is that no party will feelresponsible for the total outlays as compared to the total benefits.

However, there are also attempts to re-engineer the accountability management of joint policies andprogrammes by defining goals and evaluating results. Obviously, there are fiscal pressures to reduceoverlap and duplication between levels of government. But there are also deeper reasons that help explainthe current interest in more accountable partnerships: citizens, public managers and politicians recognisethat traditional lines of accountability are not suited for shared responsibilities and are looking for newanswers.

PUMA/RD(99)4/FINAL

12

Example

The Government of Canada redesigned federal and provincial responsibilities in the labour market sector(Canada Employment Insurance Commission, 1997: 19-26). Part II of the Employment Insurance Actpermits provinces and territories to design and deliver their own labour market programmes throughpartnership agreements with the Minister of Human Resources Development. Any provincial or territorialgovernment preferring not to assume the full responsibility for the design and the delivery of the activeemployment benefits and measures could, alternatively, choose to formalise co-management arrangements.Agreements negotiated to date reflect both these models.

Altogether, the new attention given to intergovernmental partnerships reflects the increased importance ofsub-national governments, the need to work across statutory, geographic and programme boundaries andinterest in defining and measuring performance.

Definitions and focus of the study

The focus of this study will be solely on intergovernmental partnerships. It is evident that administrativetasks may not just be reallocated between different levels of government but they may also be transferredto the private sector and the non-profit sector. As a matter of fact, the increasing importance of “public-private partnerships” and partnerships with non-governmental organisations indicates that devolution hasto be seen in a wider governance context. However, for pragmatic reasons it was necessary to limit theanalysis to partnerships between levels of government as the nature of accountability problems of private-public partnerships is somewhat different.

Partnerships among public sector institutions are of two major types (see also OECD, 1997b:34):

• Horizontal co-operation among/between different jurisdictions at the same level of governmentsuch as co-operative partnerships between municipalities.

• Vertical co-operation among/between national (federal), state (regional) and local governments.

Since the key concepts of this study -- partnerships and accountability -- are both very vague it is necessaryto provide a clear definition of these terms. In particular, partnership has become a buzzword to meananything from symbolic partnerships or performance partnership contracts to legal contracts between levelsof government. This raises the question how partnership arrangements differ from client-provider contractsin an intergovernmental context (see also Armstrong and Lenihan, 1999:13-14).

Definition of Different Contractual Arrangements

Client-Provider Contracts:

This kind of performance contract can be found between government and the private or not-for-profitsector or in hierarchical relationships within the public sector. The client specifies the tasks it wantsperformed and pays the provider for performing them. Negotiating these arrangements revolves arounddefining the terms of the contract. Managing them is about ensuring that the provider complies with theterms. In contrast to partnership agreements, client-provider contracts do not involve sharing power ordecision-making.

PUMA/RD(99)4/FINAL

13

Intergovernmental Performance Partnerships:This kind of performance contract is made between two or more autonomous parties within the publicsector. Traditionally, public sector partnerships have tended to be some sort of cost-sharing arrangements.In public sector performance partnerships, the parties jointly define performance targets for the services tobe delivered and also jointly evaluate the performance of the partnership arrangement. The management ofpublic sector performance partnerships goes beyond classical performance contracting, because it involvesresults-oriented management within a framework of shared decision-making.

Any accountability relationship -- whether in the public sector or in other social contexts --alwayspresupposes some delegated authority. In the public sector, however, accountability has a very distinctmeaning, as it is the basis for democratic societies. Which delegated authority in the public sector is held toaccount and to whom, differs from country to country and even from jurisdiction to jurisdiction (see theproceedings of the Performance Management Activity Meeting, 24-25 November 1997 on accountabilityand public organisations at http://www.oecd.org/puma/mgmtres/pac/account/ index.htm).

Yet, in all OECD Member countries, the concept of accountability has been changing and broadening overthe years from a rather narrow technical focus on financial accountability and compliance with rules andregulations to accountability for results. This new dimension of accountability does not only challengecontrol bodies such as audit courts (on the issue of “Performance Auditing”, see OECD, 1996), but also theaccountability management of different levels of governments.

It is evident that in OECD Member countries where public policies are implemented through an array ofintergovernmental arrangements, outcomes cannot be attributed to a particular level of government. Thismeans that accountability for outcomes has to be shared among different levels of government (Hatry,1997:38). A new definition has to allow for both accountability of results and for partnershipaccountability (Treasury Board Secretariat and Office of the Auditor General of Canada, 1998:4).

Definition of Shared Accountability

Accountability is a relationship based on the obligation to demonstrate and take responsibility forperformance in the light of agreed expectations. This means that in intergovernmental partnerships, thereare three kinds of accountability relationships:

• accountability among the partners;• accountability between each partner and its own governing body; and• accountability to the public.

In the paper accountability information is mainly confined to information on results (outputs andoutcomes). This does not imply that the criterion of purposefulness is imposed on the principle of legality.It rather depends on the legal and cultural framework of each OECD Member country to decide in whichareas results-oriented accountability can complement or even substitute input-oriented accountability. Withregard to the application of results-oriented accountability in Switzerland, the lawyer Mastronardi and theeconomist Schedler (1998) argue that, for many tasks of the enabling state, accountability is better securedthrough the definition of goals than through procedural rules. For example, even though all social welfarerecipients are given the same amount of welfare payments, the actual assistance they receive in terms ofresponsiveness and effectiveness may differ across jurisdictions. In this case, purely input-orientedaccountability may not secure equal treatment. However, as far as the relationships of the public

PUMA/RD(99)4/FINAL

14

administration to third parties are concerned, the principle of legality has to remain valid. The principle oflegality implies the definitions of conditions and procedures in order to protect legal equality and legalcertainty. The extent to which the legal framework emphasises the rule of law (Rechtsstaatsprinzip) differsacross countries.

The table below gives an overview of existing delegations of responsibilities and the ensuingaccountability relationship between the central government and other levels of government. The study willmainly deal with the scenarios described in lines two, three and four of table 2 where policy, managementand/or financing responsibilities are (at least formally) shared between different levels of government.Since the concept of partnership implies a certain degree of autonomy of the involved partners, each levelof government should have a say in determining the policy framework of the partnership.

PUMA/RD(99)4/FINAL

15

Table 2: Types of Delegations of Responsibility between Levels of Government

Cases Types ofDelegation

Delegation of Policy, Management andFinancing Responsibility

to Lower Levels of Government

Lines ofAccountability Example

1 Clear separation ofpolicy,management andfinancingresponsibility

Responsibility to make policy decisionsas well as to manage service delivery andto raise finance via taxes and borrowingwithin the constitutional/legislativeframework

Accountability tothe public and tothe respectivegoverning body

Devolution ofeducation policy tothe Cantonal levelin Switzerland

2 “Jointproduction”-partnerships

Shared policy, management and/orfinancing responsibility

Accountabilitymainly for outputsto the partner(s), tothe respectivegoverning bodyand to the public

The ConsortiumGreater BilbaoWater Partnershipin Spain

3 “Commonresource”-partnerships

Shared policy, management and/orfinancing responsibility

Accountabilitymainly forprocesses to thepartners(s), therespectivegoverning bodyand to the public

NationalEnvironmentalPerformancePartnership Systemin the UnitedStates

4 “Jointinvestment”-partnerships

Shared policy, management and/orfinancing responsibility

Accountabilitymainly forfinancial resourcesto the partner(s),the respectivegoverning bodyand to the public

Large cross-borderinfrastructureprojects of theEuroregion Neisse-Nisa-NysaPartnership (ERN)

5 Client-providercontract

Authority only to manage servicedelivery to meet predefined performancetargets and within the finance providedby central government

Accountability tocentral government

Next StepsAgencies in theUnited Kingdom

Source: Malcolm Morley, University College Suffolk/United Kingdom, modified by the author.

At the two extreme ends of table 2 is the case of a clear separation of policy, management and financingresponsibility at the top, and the case of a client-provider contract at the bottom. The first row reflectscases of a clear separation of political power and administrative responsibility, including the financing andspending capacity between levels of government. This scenario is currently the aim of the new fiscalequalisation system project in Switzerland (see Buschor, Hofmeister and Junod, 1999). Also Scandinaviancountries have to a large extent a clear distribution of responsibility between the central and localgovernment. It is evident that in this case no problems of blurred accountability exist. Nevertheless, as thedevolution of education policies to the Cantonal level in Switzerland shows, devolution of policies to lowerlevels of government may require a high degree of vertical and horizontal co-ordination.

PUMA/RD(99)4/FINAL

16

The fifth row at the bottom of table 2 presents the opposite scenario: the central government only transfersnarrowly defined administrative tasks within a specified financial and performance framework to anotheradministrative unit. The Next Steps Agencies in the United Kingdom are a typical example of such a client-provider relationship. In this case, the performance targets will usually be set by the client. This case is ofless interest for this study as it refers to rather classical performance contracts than to partnership contractsbetween levels of government.

The study focuses on mixed forms of delegation of responsibility which are exhibited from rows 2 to 4. Inrow 2, (at least) two levels of government share policy, management and/or financing responsibility inorder to produce goods or services together. An example of such a “joint production” partnership is theGreater Bilbao Water Partnership in Spain, where the central, regional and provincial governments fundthe operation of the municipal partnership to provide water supply and treatment services in MetropolitanBilbao. The provincial government has recently obtained voting powers in the partnership, thus equalisingthe previous mismatch between financial commitment and non-existent political rights.

Row 3 describes the case of “common resource” partnerships where different levels of government jointlymanage a common resource such as the “environment” within a shared policy framework. The NationalEnvironmental Performance Partnership System (NEPPS) in the United States (see Löffler and Parker,1999) represents such a “common resource” partnership. The U.S. Environmental Protection Agency andthe States share policy and management responsibility in order to meet jointly agreed environmental goals.

In row 4, different levels of government enter into a partnership in order to carry out an investment project.These kind of “joint investment” partnerships are quite common in the cross-border co-operation of theEuroregion Neisse-Nisa-Nysa Partnership. In this case, different levels of government from the CzechRepublic, Germany and Poland plan and implement large-scale projects with cross-border spill-over byagreeing on a common policy and financing framework (see Jerabek, Sliwa, Vidláková and Watterott,1999). The actual management of the project may be contracted out to a third party.

All three cases of intergovernmental partnerships have in common that accountability relationships arequite complex. Each partner is accountable to its respective governing body (the Parliament, the municipalcouncil etc.), to its partners and to the public. This type of partnership relationship between two or morejurisdictions is at the centre of the study.

The analytical framework

Accountability relations in partnerships are always encumbered by expectations between internal andexternal parties. The issue of accountability to whom is particularly important because different partiesmay have different expectations and performance targets according to their own intergovernmentalpartnership arrangements. In addition to this, the way and the degree to which control is carried out alsohave to be considered in accountability management.

In the public sector, four types of accountability relationship may be distinguished: hierarchical, legal,professional and political (see Radin and Romzek, 1996: 61).

Hierarchical accountability relationships are defined within public sector organisations and exhibit a highdegree of control. They are manifested in organisational roles, supervisory relationships, rules, standardsand operating procedures. The relationships are based on an expectation of obedience to organisationaldirectives.

PUMA/RD(99)4/FINAL

17

Legal accountability relationships derive from external sources that exercise a high degree of control andscrutiny. They are manifested in oversight and monitoring activities. Independent performance reviewsmay be carried out through audit courts, legislative oversight or a court review of administrative practices.

Professional accountability relationships derive from internal sources but involve a low degree of controland a high degree of discretion on how to respond to expectations of performance.

Political accountability relationships derive from external sources but involve low degrees of directcontrol. The relationships are based on an expectation of responsiveness to the stakeholders.

It follows from the above definition of shared accountability that public sector partnerships must deal withseveral sources and types of control at the same time. This means that in some instances, all four types ofaccountability relationship may be invoked simultaneously. What does this imply in terms of the selectionof a theoretical framework to analyse this web of accountability relationships in intergovernmentalpartnerships?

Given the complexities and dynamics of accountability relationships in the case of intergovernmentalpartnerships, it is obvious that principal-agent theory is in general not a very useful analytical framework.Principal-agent theory can be defined as an economic theory of co-operation that addresses issues ofdelegation and the resulting problems of control in hierarchical relationships. It is not a correct frameworkfor the analysis of accountability relationships within partnerships because principal-agent theory supposesa transfer of tasks between two partners with an unequal power relationship, whereas the concept ofpartnerships is based on the delegation of responsibility between two equal partners. In other words,principal-agent theory seems to be the appropriate analytical framework for client-provider relationshipswhere the provider (the “agent”) has no power to make policy decisions. Partnerships, however, imply thatboth parties are autonomous (see Sommermann, 1999), which suggests that both parties are “principals”. Aprincipal-agent framework is particularly problematic in the case of horizontal partnerships (for example,intermunicipal associations).

Nevertheless, principal-agent theory has the merit of highlighting imbalances of power and informationthat are very important in the context of accountability management of intergovernmental partnerships.Especially in the case of vertical intergovernmental partnerships, there will generally be an imbalance ofpower. Even though from a legal perspective all partners are autonomous, in reality there may be aprincipal-agent relationship. This may be the case even in horizontal partnerships if one of the partners ispolitically or financially more powerful than the other. Principal-agent theory also points out asymmetriesof information that may exist between the partnership institution (e.g., intergovernmental consortium) andexternal stakeholders such as the legislative bodies and the public. For example, the reporting of anintermunicipal partnership to the central government may be incomplete or biased in an effort to obtainsome form of subsidy.

Therefore the accountability management of intergovernmental partnerships will be analysed with regardto sources and types of control. When appropriate, principal-agent theory will be applied in order to focuson challenges in the implementation of performance partnerships.

The concept of performance partnerships

How can the risk of diluting or diffusing accountability in joint projects and service delivery be reduced?The concept of public sector performance partnerships suggests transforming inefficient accountabilityrelationships between external governing bodies and/or the public (the “principals”) and the partnershipinstitution (the “agent”) into accountable and efficient contractual relationships, which are based on thefollowing elements:

PUMA/RD(99)4/FINAL

18

a) clarity of roles and responsibilities, including the identification of risks;

b) joint specification of expectations of performance, including balancing expectations and capacities;

c) reporting of valid and reliable information of progress achieved against plan to the partners, thepublic and the respective governing bodies;

d) joint evaluations on how far pre-determined overall objectives have been met; and

e) feedback on the evaluation results and adjustments.

It becomes evident that the concept of performance partnerships recognises that the whole field ofaccountability is closely interconnected with the ability to set performance targets and to measure them. Ina highly political intergovernmental context, however, each of the above elements of accountabilitymanagement has its specific challenges:

a) There is a need for clarification of the roles and responsibilities of each of the parties, includingthe identification and assessment of risks during the planning stage. Since partnerships are oftenhighly experimental, they involve ample opportunities for failure. Thus, partnerships may causeunforeseen financial problems or introduce legal liabilities (see Rodal and Mulder, 1994: 32). Keystrategies for managing risks include building capacity in partners to enable them to contributefully to the partnership. For example, subnational levels of government may need managementand organisational assistance to become effective programme delivery managers.

b) In contrast to classical client-provider contracts, the concept of performance partnerships impliesthat the performance targets are not set unilaterally by the “stronger” partner. There has to be acommon understanding and agreement on the objectives being pursued and on what each party isable to contribute towards the desired goals. In the case of an unequal distribution of political andfinancial power, there is the risk that the more powerful party imposes performance targets on theother contracting party. Indeed, this behavioural problem is also one of the chief bottlenecks of theNEPPS in the United States as the States and the U.S. Environmental Protection Agency disagreeon the extent to which core performance measures are mandatory for NEPPS and non-NEPPSstates (Löffler and Parker, 1999).

c) Performance reporting in public sector partnerships does not only include the need to report to therespective governing body and to the public but also to the “partner(s)”. The information needs ofthese stakeholders may be different. A major risk is that the monitoring process becomes soabsorbing that it overshadows the primary goals of the partnership. Another difficulty of reportingin partnerships is that in general the data have to be collected from different sources. Often privacyand information protection laws only allow one specific level of government access to certain dataso that legal changes are needed in order to integrate data from different levels of government.

d) The results achieved have to be reviewed in order to consider what has been accomplished in thelight of pre-defined goals and the circumstances that existed. Truly joint evaluations presuppose acertain level of trust between the parties involved, which often still has to be developed betweendifferent the levels of government. Also staff from different levels of government usually havevery different training, skills and perceptions, which makes it a challenge to agree on somecommon evaluation methodology. Nevertheless, the implementation of joint evaluation may buildup trust and a common language, which in turn makes the next round of evaluation easier.

e) Last but not least, it is important to act upon evaluation results in order to close the accountabilitychain. The problem is that many public sector partnerships lack direct political accountability. Thismeans that even in the case of persistent low performance of a partnership, citizens have no meansto vote for a change of leadership. In most cases, there is also an absence of market competition,which could in particular improve the competitiveness of “joint production” partnerships. Ingeneral, the public is rather badly informed about the results of a partnership activity. Thus,partnerships have to become more transparent to the public. This could be achieved by consulting

PUMA/RD(99)4/FINAL

19

and involving citizens in the definition of performance indicators or by effectively communicatingthe results of the partnership activities to the public.

It is obvious that accountability is a management tool that public managers and elected officials can use incrafting and operating intergovernmental partnerships, but the success of partnerships depends on manyother factors besides accountability management. Intergovernmental relations always have a politicaldimension and in many cases, appearances may be more important for the success of the partnership thanadministrative “realities”. Nevertheless, the fragmentation of policy and service delivery across levels ofgovernment and sectors creates an increased risk of citizen alienation and distrust of governments unlessaccountability to citizens is ensured. Performance partnerships in which participants agree on performanceindicators, what targets should be established, and who is responsible for achieving the commonly agreedtargets may thus also be an important governance tool to make transparent to citizens “who is responsiblefor what”.

The following section will make the concept of performance partnerships clearer by examining differentkinds of public sector partnerships with regard to their accountability management. The comparison ofdifferent types of partnerships will show that the ability to measure and evaluate performance in terms ofoutputs and outcomes varies with the nature of the partnerships. This in turn also influences theaccountability management of a partnership. In other words, one may state that accountability problemsvary with specific forms of partnerships.

PUMA/RD(99)4/FINAL

20

PART II: ACCOUNTABILITY MANAGEMENT IN DIFFERENT KINDS OF INTERGOVERNMENTALPARTNERSHIPS

Types of intergovernmental partnerships

As mentioned above, partnerships may pursue many different purposes. The selection of case studies hasbeen guided by empirical observations about the main aspects of (vertical or horizontal) co-operativepartnerships in the public sector. Three categories of resource problems can be distinguished that give riseto project or programme partnerships (see also Lundqvist, 1998):

a) Joint production of public goods or services: These partnerships aim at the joint production ofgoods or services by making use of economies of scale. They jointly produce public or mixedgoods or services (for example, solid waste management, water treatment and publictransportation). A good example is found in public utilities which are run by two or more morejurisdictions (in general, public joint ventures).

b) Management of a common resource: This type of public sector partnership involves the jointmanagement of common resources (for example, natural resources and land use) in order to ensureresource sustainability. Individual jurisdictions might overuse the resources and create negativeexternal effects so that it is necessary to limit the usage and to establish recognised rules ofdistribution. Examples are partnerships for sustainable development, urban development andeconomic development.

c) Joint investment projects: This category includes intergovernmental partnerships whose aim is toplan and implement investment projects. The reason is that individual actors are unable to capturethe benefits from the provision of a good/service because of positive external effects. Also theresources of individual actors may be simply insufficient to provide the desired service/good at all.Classical examples of this kind of partnership are multiple-level government alliances in order toundertake infrastructure projects in the field of transportation, power plants, water projects or oiland gas extraction which are beyond the resources of any of its members.

It is evident that intergovernmental partnerships are often established to solve several resource problems atthe same time. For example, the devolution of labour market programmes from the central government tothe local level in Sweden (see Hellstrand, Joyce and Sjölander, 1999) can be considered as a “commonresource” partnership (guarantee of a national labour market) as well as a “joint production” partnership(joint provision of labour market programmes). The Swedish Ministry of Labour has made use of thispossibility as a kind of crisis management in times of high unemployment. The management of labourmarket policies in Sweden is determined by two considerations: on the one hand, the task of theemployment offices is to assist workers who cannot find a job locally to find a job in other parts of thecountry. On the other hand, jobless workers may become attractive for the local job market after adoptingthe appropriate skills profile through some local labour market programme. Local governments not onlyhave information advantages regarding the appropriate placement of the unemployed in labour marketprogrammes, but they also have a financial interest in reducing the number of young unemployed welfarerecipients.

The accountability management of labour market programmes with municipal co-management is based onLocal Employment Office Boards that are supposed to co-ordinate and supervise the activities of the state-run local employment offices and the municipalities. Since 1996, the municipalities have the right toappoint the majority of the board members. This highly problematic management structure, as well as thefiscal incentives of Swedish municipalities to keep as many residents (even unemployed) as possible in

PUMA/RD(99)4/FINAL

21

their constituency and the high volume demand of the central government for labour market programmes,is likely to induce the Local Employment Office Boards to hamper mobility of the unemployed. Because ofthis imminent danger, the State labour market administration has kept the right to override the decisions ofthe Local Employment Office Board. This shows that the existence of different objectives within a publicsector partnership will inevitably cause accountability problems.

Nevertheless, for analytical purposes, it seems to be legitimate to focus on the main aim of eachpartnership. This does not exclude attention also being given to accountability problems resulting from theco-existence of several (competing) objectives within an intergovernmental partnership.

As the three different types of intergovernmental partnerships illustrate, partnerships deal with differentkinds of public goods and services. Theoretical literature on performance management (see, Wilson 1989)suggests that the type of the public good or service strongly influences the technical ability to measureperformance (in terms of outputs) and to carry out evaluations (in terms of outcomes). As a result,partnerships often measure what is easily measurable but not necessarily what should be measured from anaccountability point of view. Even though this a generic public management problem, outputs andoutcomes which are not measured in partnerships gives the members much more leeway to bluraccountability than is the case in arrangements with less complex lines of accountability.

As a result, different types of partnerships can be expected to have different kinds of challenges foraccountability management:

• “Joint production” partnerships focus on the provision of “hard” or technical services such asservices to property and to general infrastructure. These are services with a relatively highproportion of fixed capital costs, compared to variable (personnel) costs, which results indecreasing average unit costs if the output is increased. In general, output monitoring of “jointproduction” partnerships will be technically easy provided that commonly agreed output targetsexist. However, since the impact of the production activity is beyond the control of any member,there are little incentives to evaluate the overall performance of the “joint production” partnershipin terms of outcomes.

• “Common resource” partnerships deal with the use of common goods such as land and naturalresources. Due to their regulatory nature, the management focus is on processes rather than onresults. This means that the performance of such partnerships are usually evaluated in terms ofcompliance with explicit rules. However, process management is not a purpose in itself but onlyserves to realise medium and long-term outcomes. Therefore, meaningful performance indicatorshave to be outcome-based. The problem is that outcomes are often only visible in the long-termand cause-effect relationships are unclear. As a result, “common resource” partnerships experiencedifficulties to shift the focus from processes to outcomes.

• “Joint investment” partnerships involve long-term planning but also the raising of a considerableamount of financial resources. Due to the long time frames involved in major project development,politicians have an incentive to develop over-optimistic forecasts of project viability in order tohave projects approved. But they are often no longer in office when actual viability can becalculated. As a result, ex-post evaluations of the actual viability of large infrastructure projects areoften missing completely. Also, project development is typically seen as a technical exercise withan early focus on technical solutions, whereas accountability issues relating to the implementationand evaluation of proposed projects play a minor role in project preparation. As a result, “jointinvestment” partnerships tend to focus mostly on the “who pays how much” question.

PUMA/RD(99)4/FINAL

22

The matrix below summarises the dominant performance orientation in different kinds ofintergovernmental partnerships. It becomes evident that the evaluation of outcomes is the greatestchallenge for all types of partnerships.

Table 3: Performance orientations in intergovernmental partnerships

Type ofPartnership

Dominant focus oninputs (financial

resources)

Dominant focus onprocesses

Dominant focus onoutputs

Dominant focus onoutcomes

“JointProduction”

X

“CommonResource”

X

“JointInvestment”

X

Source: Arthur Benz, University of Halle-Wittenberg/Germany, completed and modified by the author.

The difficulty in evaluating outcomes but also in measuring outputs for certain types of partnerships willalso influence the management of accountability. Questions arise: what does this imply for publicmanagers? How can the accountability management of these different types of partnerships be improved?The next chapter will point out how contractual arrangements may support the accountability managementof the various partnerships. This will be done by drawing from the partnership case studies.

Accountability management of “joint production” partnerships

“Joint production” partnerships are characterised by the co-operation of two or more jurisdictions in orderto produce goods or services together. This usually involves the establishment of some kind of commoninstitution which acts on behalf of the “principals”.

As “joint production” partnerships are output-oriented, the specification of output targets is relatively easy,especially since joint production partnerships will in general focus on industrial services. It is evident thatin the case of personalised services, economies of scale are much harder to achieve. This implies thatoutput-oriented performance information can be expected to be readily available. However, outcomeevaluations of the “joint production” partnership by the owner governments as well as the use of effective(positive or negative) sanctions are often missing due to inappropriate accountability structures.

An example of “joint production” partnerships are the mancomunidades in Spain, which are municipalmulti-purpose associations (with possible participation by the third sector). Especially small municipalitiesin rural areas use this intergovernmental arrangement for the purpose of joint service delivery. The Uribe-Kosta Services Partnership in Metropolitan Bilbao illustrates very well the accountability problems arisingfrom traditional agency relationships (see Font, Gutiérrez Suárez and Parrado Díez, 1999).

The Uribe-Kosta Services Partnership (UKSP) delivers a wide range of services, including wastemanagement, social services, consumer information and protection of the Basque language. Over time, theUKSP has assumed more and more municipal tasks. The great variety and the nature of the services

PUMA/RD(99)4/FINAL

23

provided by the UKSP have two negative implications for accountability management: first, electedpolitical leaders have a great interest in the affordability of social and cultural services for their electorate.Therefore, they tend to fix the price of “politically important” services below the average costs estimated,which means that higher levels of government have to fund part of the running costs. Secondly, themanagers of the UKSP do not have a great deal of autonomy, as municipal councillors have strongincentives to micro-manage the UKSP. The wide spectrum and the non-technical nature of the servicesgives them many opportunities to interfere with operational management. Mancomunidades even give riseto financial and legal accountability problems as members try to avoid making their contributions. Therecurrent problem with the financial commitment of municipal members in mancomunidades made theProvince of Biscay pass a special law in order to punish “free-rider behaviour”.

Accountability management of “joint production”-partnerships

Weaknesses:

In general, the owner governments will have a strong political interest in the services provided by the“joint production” partnership so managerial accountability will tend to be weak. In most cases, thegoverning body of the partnership is appointed by the political majorities of the member jurisdictions, thuspossibly giving rise to a conflict of interests. In addition to this, the management of the partnership tends tohave only little decision-making autonomy, which often goes hand in hand with insufficient professionalstaff. All in all, weak managerial accountability of the joint institution and “free-rider behaviour” of theowner governments will result in inefficient production.

Recommendations:

In order to ensure more efficient service provision by the partnership, a competitive environment must becreated (ranging from market competition through managed competition, to forms of quasi-competitionlike benchmarking). Performance contracts between the owner governments and the governing board mustensure that the board has incentives to evaluate the overall performance of the partnership. There also haveto be professional managers with a sufficient degree of decision-making autonomy and performance-basedemployment contracts. The establishment of financial accounting and management accounting systemsmay enable the governing body to monitor the managers’ performance. The involvement of citizens in thedefinition of performance targets may also increase the transparency of “joint production” partnerships andthus accountability to the public.

Even though none of the partnership case studies represents this kind of “contractual productionpartnership” in pure form, the Greater Bilbao Water Partnership can be considered as a promising goodpractice (see Font, Gutiérrez Suárez and Parrado Díez, 1999). Consortia are one-purpose associationsbetween municipalities and higher levels of government, often with the involvement of the private sector.Most types of consortia have a professional manager who is accountable to the politically appointedmembers of the consortium’s governing body. Because of the technical nature of the service provided byconsortia, politicians have less incentives to micro-manage the partnership.

The Greater Bilbao Water Partnership represents a vertical intergovernmental partnership with a highlydecentralised structure of control. The 19 founding municipalities and the other 24 municipalities whichjoined the partnership later have voting powers to make strategic decisions as well as executive powers tomonitor the performance of the managers of the GBWP. The managers provide the governing board withtimely financial information as well as information on outputs. Outcome-based information is weak due tothe dispersion of administrative responsibilities in this field. Since the provincial government is the mostimportant financial contributor to the partnership, it uses its financial power to determine the overall policyand to sanction bad performance of the consortium’s management. Clearly, a better solution would be to

PUMA/RD(99)4/FINAL

24

make explicit from the beginning the financing intended for current expenditures and for investments.Contractual arrangements might also include performance-based employment contracts with theconsortium’s management, including the possibility of contract termination in the case of continued badeconomic performance of the consortium.

Accountability management of “common resource” partnerships

This form of intergovernmental arrangement is intended to manage jointly a particular common resource.The most typical examples are land use planning, environmental protection and the management of otherpublic goods. Due to devolution, this kind of public sector partnership is becoming more and morecommon. This means that higher levels of government transfer responsibility to lower levels ofgovernment in order to allow for a better match between “regional” preferences and national demands.

In Switzerland, the Swiss Maturity Board which deals with the recognition of certificates of maturity canbe regarded as a “common resource” partnership of education programmes. As the Swiss case study on thedevolution of educational policies to the Cantonal level shows, Cantons have always enjoyed largeautonomy in educational policies (see Buschor, Hofmeister and Junod, 1999). This implies that educationpolicies can be quite different from Canton to Canton: for example, in the Canton of Zürich, the primarytarget for educational services is economic efficiency (costs-per-student), whereas in the Canton of Genevathe focus is on user satisfaction (students and parents). Nevertheless, the resulting diversity is notevaluated. This implies that the Swiss Maturity Board has to focus on the fulfillment of proceduralconditions to be fulfilled by grammar schools instead of qualitative and quantitative performancestandards. In the end, certificates from grammar schools in each of the 26 Cantons that fulfil the provisionsof the Swiss Maturity Board may be legally equal. The unanswered question is whether they are equal ineducational quality.

Accountability management of “common resource” partnerships

Weaknesses:Because of the process-oriented nature of the partnership, the performance orientation is low. This makes itvery difficult for the “owner” governments to evaluate whether the common resource is managed in aneffective way. As a result, compliance with processes often becomes an end in itself.

Recommendations:“Common resource” partnerships must strengthen their performance orientation by identifying meaningfuloutcome indicators. The challenge is to find the appropriate mix between procedural requirements andoutcome targets. The definition of outcome indicators also offers a chance to consult with citizens and tostrengthen accountability relationships with the public by reporting publicly progress against pre-determined outcome targets.

The achievements but also the difficulties with the creation of the National Environmental PerformancePartnership System (NEPPS) illustrate very well the transition from classical “common resource”-partnerships with high costs of control and little performance orientation to a more performance-basedmanagement of common resources (see Löffler and Parker, 1999).

The system was designed to improve the quality of the environment across the United States and, at thesame time, to allow the States to deal with local problems through innovative approaches. The basic idea ofenvironmental performance agreements between the States and the U.S. Environmental Protection Agency

PUMA/RD(99)4/FINAL

25

(EPA) was to trade off reduced reporting to the federal level for a set of core performance indicators forstate-level environmental management. The definition and interpretation of those core performanceindicators proved to be the greatest difficulty: whereas the States wanted to align the performanceindicators with their reporting systems, the national level wanted to integrate them into the GovernmentPerformance and Results Act (GPRA). Also, EPA was reluctant in many cases to reduce federal control.This meant that the incentives provided by the performance agreements were inadequate. The situation hasbeen aggravated further by the fact that the performance leadership programme, which was meant to offeradditional incentives to high-performing States, has not been implemented yet.

Accountability management of “joint investment” partnerships

This type of intergovernmental partnership will usually be redistributive in nature, as the pooling ofresources involves the alignment of jurisdictions with a weak resource base with financially strongjurisdictions. In general this implies the involvement of higher levels of government. The partners tend toadminister the resources through a separate fund with its own administrative structure. In the case of largepooling, the administration of a common fund may require the establishment of international organisations.This kind of “international partnership” has very complex multi-level accountability structures. Onereason is the multitude of actors involved in such partnerships, which are often embedded in differentadministrative and political structures. Another reason is the necessity to establish a dual accountabilitystructure: accountability relationships for providing the resources and for using them.

“Joint investment” partnerships will usually have great difficulties in the specification and evaluation ofperformance. The problem is that the main aim of “joint investment” partnerships is to raise funds toexecute rather broadly defined infrastructure projects. Also the management of performance information isa challenge in itself. Because of the multitude of stakeholders involved, the performance data are notalways produced in a timely manner nor channelled to the appropriate user. Evaluation of results aremethodologically and politically difficult because the objectives of the investment project have not beenmade explicit in terms of unambiguous performance targets.

The Euroregion Neisse-Nisa-Nysa (ERN), which is a municipal partnership of three border territoriesbetween Germany, Poland and the Czech Republic (see Jerabek, Sliwa, Vidláková and Watterott, 1999), isthe organisational framework for many kinds of horizontal, vertical and in particular mixed forms of cross-border partnerships. Because of the uneven economic and financial power of the municipal memberassociations on the Polish and Czech side on the one hand, and on the German side on the other hand, thepartnership has a strong redistributive character.

Different administrative structures, language problems, the divergence of economic development and ofnational interests are the daily obstacles for municipal co-operation. The ERN also has to cope with thedifficulty of having to rely on unprofessional working groups to a large degree. Altogether, insufficientstaffing of the ERN bodies and the home secretariats make it impossible to manage the huge informationflow that is involved with cross-border project work and to evaluate the results of joint projects at least ona periodic basis.

The weak financial basis of the ERN makes it dependent on transfers from third parties. Especially on thePolish side, the EU Phare Cross-Border Co-operation Programme uses the organisational and managementstructure of ERN to carry out small-scale projects. Even though the EU Programme has ex-ante and ex-post project evaluation criteria, there are all kinds of performance evaluation problems, makingaccountability management difficult. For example, the measurement of the cross-border impacts can yieldvery different results, depending on which stakeholder is responsible for the evaluation. Also, projectproposals are often made on the basis of insufficient information and unrealistic expectations concerning

PUMA/RD(99)4/FINAL

26

the size of the funding. But the EU Programme provides for some positive incentives for successfulprojects by granting a financial reward of 5 per cent of the transfer payment to the ERN.

In spite of the complex accountability structure of the ERN, the cross-border partnership has been verysuccessful in solving cross-border problems of a municipal scale. This includes “joint production”partnerships to jointly manage common sewage systems as well as “common resource” partnerships toextend bicycle paths and other tourist facilities across the border. However, as surveys reveal, the ERN isless effective in dealing with cross-border problems of a regional scope. The involvement of the ERN inregional projects also risks the weakening of accountability to the local authorities. Regional policiesshould be transferred to the regional administrative level in the three countries, provided that such a levelwill exist in the Czech Republic by the year 2000.

Accountability management of “joint investment” partnerships

Weaknesses:

Performance specifications regarding the outputs and outcomes to be achieved through the investmentproject are often missing. The seemingly “free” resources may lead to inefficient use by potential externalbeneficiaries. In the absence of institutional arrangements that promote accountability in the projectplanning and development phase, risk identification and allocation between the partners will be lacking aswell.

Recommendations:There is also a need to formulate clear and credible performance criteria already in the project planningphase. There is also a need to take into account the risks that are associated with long-term investmentprojects and to allocate responsibility for carrying the costs of uncertainty. Performance contracts betweenthe partnership institution and potential external beneficiaries of investment projects may establish anaccountability link by requesting beneficiaries to contribute substantially to the amount of transfers.

Shared transfer programmes are a classical example of “joint investment project” partnerships with highagency costs. These efficiency losses have induced the Swiss Government as well as the Government ofthe Region of Cantabria in Spain to reform intergovernmental transfers through contract management.

As the Swiss case study explains, the ongoing reform of the federal financial equalisation system does notonly pursue fiscal goals but also aims at a restructuring of responsibility between levels of government (seeBuschor, Hofmeister and Junod, 1999). The overall objective is to re-establish the principle of equivalencefor as many tasks as possible. This means assigning spending powers and taxing powers to thegovernmental level where the task is carried out. Where institutionalised co-operation between levels ofgovernment is inevitable, performance contracts are defined that are supposed to reduce the agency costsfrom intergovernmental transfers.

To sum up, public sector partnerships that are based on performance contracts are clearly in a betterposition to reduce the risk of blurred accountability than merely cost-sharing arrangements betweendifferent jurisdictions. Yet, the case studies show that the application of contract management is still in anearly stage between different levels of government. This is not surprising given the technical and politicaldifficulties involved in specifying and evaluating performance of intergovernmental transactions. It isobvious that less specified performance partnerships require a high degree of trust between political andpublic managers who represent different constituencies and potentially different interests. This raises thequestion of the degree to which intergovernmental performance partnerships must be legally enforceable.

PUMA/RD(99)4/FINAL

27

PART III: PERSPECTIVES AND LIMITS OF INTERGOVERNMENTAL PERFORMANCE PARTNERSHIPS

Legal contracts and non-statutory contracts mainly differ with regard to their specification and disputesettlement mechanisms: whereas legal contracts can be taken to court, there is less legal protection ofinformal contracts (which, however, does not imply that they are not enforceable). Legal contracts attemptto provide for as many future contingencies as can be foreseen, or to agree on compensation payments ex-ante in case the contract cannot be fulfilled. By doing so, legal contracts allow for a credible threat of(relatively) low-cost legal settlement. Of course, this is not possible in a world of incomplete informationand limited rationality, so in reality legal contracts are not fully specified and require some degree of trust.

Informal contracts, by contrast, are completely based on trust with regard to endogenous risks (given theirlack of judicial enforceability). Their logic is very different: they do not specify the set of economicallyrelevant contingencies but rather rely on risk absorption through the establishment of adequate incentivestructures. As a result, informal contracts would invoke high-cost legal settlement in the case of disputes.This means that informal contracts between levels of government are only viable if a political andadministrative system has developed social incentives for the protection of trust. Is such a culture of trustrealistic in a highly dynamic environment with a multitude of actors? The diffusion of trust in anintergovernmental context presupposes a critical mass of organisations that already work on the basis oftrust relationships as opposed to control and sanctions. The probability that such a critical mass can beobtained can be expected to vary with the complexity of the political system. In a country where thenumber of decision-makers in the public sector domain is rather small and stable it is easier to build uptrust relationships than in a big country with multiple levels of government where the pool of publicdecision-makers is bigger and more subject to fluctuations.

Nevertheless, the high turnover rate of political decision-makers due to the co-existence of several electoralcycles is not beneficial for the establishment of informal performance partnerships. Trust relationships willonly pay off in the long term and may require high transaction costs in the initial stage. The NEPPSpartnerships offer a good example for the difficulties in building up this trust capital. Of course, long-termgains are very problematic incentives in a political world where actors are guided by short-term legislativeperiods. Especially in the case of horizontal public sector partnerships with more or less equal partners, theactors involved may find legal contracts preferable to non-statutory contracts. Whereas the “subordination”relationships in vertical partnerships offer the pragmatic solution of using the political or financial powersof the central government to sanction non-fulfilment of partnership obligations, horizontal partnerships aremore prone to “free-rider behaviour”. Last but not least, intergovernmental partnerships must fit into thenational legal order (see Sommermann, 1999). In many OECD Member countries, different levels ofgovernment may conclude agreements with each other which are judicially enforceable.

This does not exclude the complementary use of legal contracts and non-statutory performance contracts.For example, transfer payments may be accorded to a lower level of government on the basis of a unilaterallegal provision which only specifies the amount of the grant based on certain objective criteria. Theperformance targets to be achieved through the transfer payment would be jointly specified in an informalperformance contract. The other trend that is becoming more and more common is to change the nature ofthe administrative law, which means granting subsidies on the basis of administrative contracts rather thanadministrative acts. So far, “performance partnerships” between levels of government are still treated asexceptions in the legal framework, as shown by the use of waivers in the United States. Asintergovernmental performance partnerships become more common, the legal framework will have to bechanged accordingly.

PUMA/RD(99)4/FINAL

28

REFERENCES

ARMSTRONG, Jim and LENIHAN, Donald G. (1999), “From Controlling to Collaborating: WhenGovernments Want to be Partners. A Report on the Collaborative Partnership Project”, New Directions,No. 3, Institute of Public Administration of Canada: Toronto.

BOORSMA, Peter B. and DE VRIES, Piet (1998), “Relative Efficiency of Intermunicipal Provision”,in: Halachmi, Arie and Boorsma, Peter B. (eds.), Inter and Intra Government Arrangements forProductivity. An Agency Approach, Kluwer Academic Publishers: Boston, Dordrecht and London, pp. 137-145.

BRUZELIUS, Nils; FLYVBJERG, Bent and ROTHENGATTER, Werner (1998), “Big Decisions, BigRisks: Improving Accountability in Mega Projects”, International Review of Administrative Sciences, Vol.64, pp. 423-440.

BUSCHOR, Ernst; HOFMEISTER, Albert and JUNOD, Raymond (1999), The Devolution of EducationPolicies from the Federal Government to the Cantons in Switzerland, , available on internet:http//:www.oecd.org/puma/

CANADA EMPLOYMENT INSURANCE COMMISSION (1997), 1997 Employment Insurance.Monitoring and Assessment Report, Ministry of Human Resources Development: Ottawa.

FONT, Javier; GUTIÉRREZ SUÁREZ, Rafael and PARRADO DÍEZ, Salvador (1999), Inter-governmental Partnerships at the Local Level in Spain: Mancomunidades and Consortia in a ComparativePerspective, available on internet: http//:www.oecd.org/puma/

FURUKAWA, Shun’ichi (1998), “Federalizing Japan: A Silent Revolution?”, Institute of Policy andPlanning Sciences. Discussion Paper Series No. 774, University of Tsukuba, Ibaraki, Japan.

GESCHÄFTSSTELLE DES SACHVERSTÄNDIGENRATS “SCHLANKER STAAT” (1997), Lean StateAdvisory Committee. Final Report, Bonn (internal document).

HATRY, Harry P. (1997), “We Need a New Concept of Accountability”, The Public Manager, fall,p. 37f.

HELLSTRAND, Håkan; JOYCE, Patrick and SJÖLANDER, Katarina (1999), The Devolution of LabourMarket Programmes from the Central Government to the Local Level in Sweden, available on internet:http//:www.oecd.org/puma/.

JERÁBEK, Milan; SLIWA, Tomasz; VIDLÁKOVÁ, Olga and WATTEROTT, Gerhard (1999),Administrative Co-operation in the Euroregion Neisse-Nisa-Nysa, available on internet:http//:www.oecd.org/puma/

LÖFFLER, Elke and PARKER, Christina (1999), The National Environmental Performance PartnershipSystem (NEPPS) between the States and the U.S. Environmental Protection Agency, available on internet:http//:www.oecd.org/puma/

LUNDQVIST, Lennart J. (1998), “Local-to-Local Partnerships among Swedish Municipalities: Why andHow Neighbours Join to Alleviate Resource Constraints”, in: Pierre, Jon (ed.), Partnerships in UrbanGovernance. European and American Experience, MacMillan Press: Houndsmills and London, pp. 93-111.

MASTRONARDI, Philippe and SCHEDLER, Kuno (1998), New Public Management in Staat und Recht,Haupt: Bern, Stuttgart and Vienna.

OECD (1996), Performance Auditing and the Modernisation of Government, Paris.

PUMA/RD(99)4/FINAL

29

OECD (1997a), Managing Across Levels of Government, available on internet:http//:www.oecd.org/puma/

OECD (1997b), Partnership in the United States. OECD Reviews of Rural Policy, Paris.

RADIN, Beryl A. and ROMZEK, Barbara (1996), “Accountability Expectations in an IntergovernmentalArena: The National Rural Development Partnership”, Publius: The Journal of Federalism, Vol. 26, No. 2,pp. 59-81.

RODAL, Alti, and MULDER, Nick (1994), “Partnerships, Devolution and Power-Sharing: Issues andImplications for Management”, Optimum. The Journal of Public Sector Management, Vol. 24, No. 3(winter), pp. 27-48.

SOMMERMANN, Karl-Peter (1999), Accountability Management of Intergovernmental Partnerships in aLegal Perspective, available on internet: http//:www.oecd.org/puma/

TREASURY BOARD SECRETARIAT and OFFICE OF THE AUDITOR GENERAL OF CANADA(1998), Modernizing Accountability Practices in the Public Sector (http://www.tbs-sct.gc.ca/rma/account/OAGTBS_E.html)

WILSON, James Q. (1989), Bureaucracy: What Government Agencies Do And Why They Do It, BasicBooks: New York.

PUMA/RD(99)4/FINAL

30

Notes

1 . “The author would like to thank Jim Brumby, OECD Public Management Service, Klaus Lüder, German Post-Graduate School of Administrative Sciences in Speyer, John Mayne, Office of the Auditor General of Canada,Malcolm Morley, University College Suffolk, United Kingdom and Beryl A. Radin, University of Albany-StateUniversity of New York for their inputs to this paper”.

2. Devolution has been used in other work by the Public Management Service as an umbrella term covering allforms of transfers of responsibility (for definitions of decentralisation and deconcentration, see OECD,1997a:18).